Abstract: In view of the large number of electronic commerce returns, a kind of integrated distribution mode is put forward, which is based on the third party logistics and self built logistics. Location Routing Problem （LRP） has become a very difficult problem to be solved. In order to reduce distribution costs and improve customer satisfaction, using the improved genetic algorithm, the first time window is added in the self built logistics cost penalty function, remote area distribution is used third party logistics distribution mode, pick up reverse logistics are established as one of the current location of the path model; then aiming at the problem of the complexity, a genetic algorithm is designed to solve the model, the time distribution thought is used for initialization, local search algorithm is introduced, adaptive genetic algorithm is improved; finally, through data analysis for reducing logistics costs, reduce the running time, improving the operating efficiency which verifies the validity of the model feasibility and algorithm.

Key words: reverse logistics, third party distribution, e-commerce, location route

摘要： 针对电子商务退货的大量产生，提出一种送取一体化的配送方式，第三方物流和自建物流结合的逆向物流配送策略。选址和路径问题LRP（Location Routing Problem）成为了很多物流配送企业正在面对和急需解决的难题。为了降低配送方式成本以及提高客户满意度，利用改进的遗传算法，首先在自建物流成本中加入时间窗惩罚函数，偏远地区的配送采用第三方物流配送方式，建立送取货为一体的逆向物流选址路径模型；然后针对问题的复杂性设计一种改进的遗传算法对模型进行求解，采用时间分布思想初始化种群，引入局部搜索算法，设计自适应的改进遗传算法；最后采用实验进行数据分析，通过物流成本的降低、运行时间的减少、运行效率的提高等数据验证了模型的可行性以及算法的有效性。

关键词: 逆向物流, 第三方配送, 电子商务, 选址路径